AWS Compute: The Work Chooses the Compute
“The wand chooses the wizard, Mr. Potter.” — Ollivander
The Business Goal
Back to Bits Boba
The Compute section began with one question: what kind of work needs to run?
Some work needed a permanent shop with full control: EC2. Some needed identical shops behind one host: AMIs, Auto Scaling, and Elastic Load Balancing. Some needed portable kitchens: containers, ECS, Fargate, ECR, and EKS. Some needed a catering company to finish a finite backlog: Batch. And some needed a runner for one bounded errand: Lambda.
The runner became more real as the story continued. A custom boba order could wait at the counter, receive an order number, or sit on a ticket rail. A spill needed retries, an owned recovery shelf, and idempotent handling. The runner needed a kit, a clearance badge, and a safe shift change for a new recipe.
The Decision Was Never “Serverless or Not?”
The question was never which service sounded most modern.
What shape is the work?
| Work says… | Compute answer |
|---|---|
| “I need a machine I control.” | Amazon EC2 |
| “I need a fleet of identical application instances.” | EC2 + Auto Scaling + ELB |
| “I need a packaged application runtime.” | Containers + ECS/EKS/Fargate |
| “I need a finite backlog to finish.” | AWS Batch |
| “I react to an event and finish within a boundary.” | AWS Lambda |
The Last Bite
You did not just learn a list of compute services. You learned to listen to the work before choosing the kitchen.
The work chooses the compute.
Next section: AWS Storage: The Memory Chooses Its Home
Compute explains where work runs. The next Services section explains where its files, images, artifacts, and durable object data live. Architecture will become its own later top-level collection, where the series can compare monoliths, modular monoliths, request/response paths, events, orchestration, and choreography across the completed service stories.